Home Dec 2026 Elections CepRass Snapshot Favors Barrow—Not a Forecast, Statistician Warns

CepRass Snapshot Favors Barrow—Not a Forecast, Statistician Warns

0
33
Dr. Ebrima Bayo, a Gambian-born US-based statistician and mathematician.

President Adama Barrow holds a clear lead in the latest CepRass opinion poll, but a detailed statistical reading of the figures shows the race remains fluid with just weeks to go before The Gambia’s 5 December presidential election, according to an independent analysis by Dr. Ebrima Bayo, a Gambian-born US-based statistician and mathematician.

The Center for Policy, Research and Strategic Studies (CepRass) survey, conducted face-to-face from 17 October to 1 November among 1,180 respondents holding valid voter cards across 118 enumeration areas, put the NPP-led alliance headed by Barrow at 37.9 percent, the Gambia First Coalition led by Ousainou Darboe at 24.2 percent, and Talib Ahmed Bensouda’s Coalition 2026 at 11.3 percent. Smaller shares went to the No-To-Alliance (5.8 percent) and other groups, with 11.6 percent undecided and 4.2 percent preferring not to disclose their preference.

Bayo, who has a background in mathematics, statistics and economics, stressed that an opinion poll is a snapshot of stated intentions, not an election result. Barrow’s 13.7-point advantage over Darboe is substantial, he noted, yet 62.1 percent of respondents did not choose the incumbent. Widespread economic dissatisfaction—80 percent rating the national economy poorly and 52 percent calling government delivery on promises ineffective or very ineffective—has not automatically translated into opposition support, suggesting competence and other factors continue to sustain Barrow’s position.

Closing the gap is mathematically demanding. A direct switch of 6.85 percentage points of respondents from Barrow to Darboe would erase the lead, Bayo calculated, because each such voter reduces one candidate’s total while increasing the other’s. Gaining the same number solely from undecided voters would require more than 13.7 points. The source of any new support therefore matters as much as the volume.

Bensouda’s 11.3 percent draws attention as a potential kingmaker. Combining Darboe and Bensouda yields 35.5 percent; adding the No-To-Alliance’s 5.8 percent produces 41.3 percent, 3.4 points above Barrow. But alliances are not simple addition. If only 80 percent of those supporters stayed with a merged ticket, the combined figure would fall to 28.4 percent—well below Barrow. Retention and the ability to attract additional voters, especially among younger respondents where Bensouda reportedly performs better, will determine whether his base expands or shifts.

The 15.8 percent of undecided and non-disclosing respondents exceeds the Barrow-Darboe gap. In an extreme illustration, if all moved to Darboe while Barrow’s support stayed constant, Darboe would reach 40 percent. Real voters are unlikely to move as a bloc, yet the size of this group underscores why it cannot be ignored, particularly in populous urban areas such as Kanifing, Banjul and Brikama where undecided rates are higher.

Statistical caveats further temper interpretation. Under simple random sampling, the approximate maximum margin of error at 95 percent confidence would be about ±2.85 percent, but CepRass used multistage stratified sampling, so design effects, clustering, and weighting are needed for precise uncertainty estimates. Household heads answered the voting-intention question in 77.4 percent of interviews, raising possible selection bias if they substituted for originally selected respondents. Coverage of all constituencies does not guarantee equally reliable local estimates. The margin of error addresses only sampling variation, not nonresponse, inaccurate reporting or turnout. Without comparable earlier polls, trends cannot be firmly established.

Bayo concluded that Barrow is clearly leading among those surveyed, Darboe remains the closest challenger, and Bensouda is a significant third force. For the incumbent, the question is whether his 37.9 percent will hold if opposition support consolidates. For Darboe, the arithmetic favors converting Barrow voters over purely attracting undecided ones. For Bensouda, the issue is whether 11.3 percent can grow or serve as a transferable base. Smaller parties’ influence depends on how their supporters behave under different scenarios.

He welcomed CepRass’s contribution to public debate but called for fuller publication of the methodology, including weighting, design effect, constituency sample sizes, and respondent selection. The poll, he wrote, shows who is ahead among those interviewed but does not determine who will win. An estimate is not a certainty, and voter behavior is not fixed.

LEAVE A REPLY

Please enter your comment!
Please enter your name here